U.S. patent application number 11/147303 was filed with the patent office on 2006-12-14 for security perimeters.
This patent application is currently assigned to MCI, Inc.. Invention is credited to Douglas M. Pasko.
Application Number | 20060282891 11/147303 |
Document ID | / |
Family ID | 37499106 |
Filed Date | 2006-12-14 |
United States Patent
Application |
20060282891 |
Kind Code |
A1 |
Pasko; Douglas M. |
December 14, 2006 |
Security perimeters
Abstract
A security system that is associated with a customer network
includes first, second, and third security perimeters. The first
security perimeter includes a set of content delivery network (CDN)
devices configured to provide first protection against a network
attack associated with the customer network. The second security
perimeter includes a set of mitigation devices configured to
provide second protection in terms of mitigation services as a
result of a network attack associated with the customer network.
The third security perimeter includes a set of hierarchy devices
configured to provide third protection against a network attack
associated with the customer network.
Inventors: |
Pasko; Douglas M.; (Herndon,
VA) |
Correspondence
Address: |
VERIZON;PATENT MANAGEMENT GROUP
1515 N. COURTHOUSE ROAD
SUITE 500
ARLINGTON
VA
22201-2909
US
|
Assignee: |
MCI, Inc.
Ashburn
VA
|
Family ID: |
37499106 |
Appl. No.: |
11/147303 |
Filed: |
June 8, 2005 |
Current U.S.
Class: |
726/23 |
Current CPC
Class: |
H04L 63/0209 20130101;
H04L 63/1458 20130101 |
Class at
Publication: |
726/023 |
International
Class: |
G06F 12/14 20060101
G06F012/14 |
Claims
1. A security system associated with a customer network,
comprising: a first security perimeter including a plurality of
content delivery network (CDN) devices configured to provide first
protection against a network attack associated with the customer
network; a second security perimeter including a plurality of
mitigation devices configured to provide second protection in terms
of mitigation services as a result of a network attack associated
with the customer network; and a third security perimeter including
a plurality of hierarchy devices configured to provide third
protection against a network attack associated with the customer
network.
2. The security system of claim 1, wherein the first, second, and
third security perimeters are located outside of the customer
network.
3. The security system of claim 1, wherein each of the CDN devices
is configured to act as a termination point for traffic associated
with the customer network.
4. The security system of claim 1, wherein each of the CDN devices
is configured to serve replicated content associated with the
customer network and provide security services with regard to
requests for the replicated content.
5. The security system of claim 4, wherein each of the CDN devices
is further configured to obscure the presence of the customer
network.
6. The security system of claim 1, wherein one of the mitigation
devices is configured to: receive traffic intended for the customer
network, process the received traffic to remove malicious traffic
and leave legitimate traffic, and forward the legitimate traffic
toward the customer network.
7. The security system of claim 6, wherein a tunnel is configured
between the one of the mitigation devices and one of the hierarchy
devices; and wherein when forwarding the legitimate traffic toward
the customer network, the one of the mitigation devices is
configured to: apply a tunnel encapsulation to the legitimate
traffic, and forward the legitimate traffic with the tunnel
encapsulation to the one of the hierarchy devices via the
tunnel.
8. The security system of claim 1, wherein each of the hierarchy
devices is configured to act as a gateway for traffic intended for
the customer network to obscure the presence of the customer
network.
9. The security system of claim 1, wherein one of the hierarchy
devices is configured to determine when the customer network is the
subject of a network attack; and wherein one of the mitigation
devices is configured to receive traffic redirected from the
customer network when the customer network is the subject of the
network attack.
10. The security system of claim 1, wherein the hierarchy devices
are configured to receive traffic from the CDN devices and the
mitigation devices.
11. The security system of claim 1, wherein the CDN devices and the
hierarchy devices are configured to obscure the presence of the
customer network to protect the customer network against network
attacks during system patching or upgrading within the customer
network.
12. The security system of claim 1, wherein the CDN devices and the
hierarchy devices are configured to use network attack detection or
prevention techniques to identify or process malicious traffic
intended for the customer network.
13. A security system associated with a customer network,
comprising: a first security perimeter including a plurality of
content delivery network (CDN) devices geographically distributed
in a public network, the CDN devices being configured to serve as a
first termination point for a first set of traffic intended for the
customer network; and a second security perimeter including a
plurality of hierarchy devices located within the public network,
the hierarchy devices being configured to serve as a second
termination point for a second set of traffic intended for the
customer network, where the second set of traffic includes traffic
from the CDN devices and traffic different from the traffic from
the CDN devices.
14. The security system of claim 13, wherein the CDN devices and
the hierarchy devices are configured to obscure the presence of the
customer network to protect the customer network against network
attacks during system patching or upgrading within the customer
network.
15. The security system of claim 13, wherein the CDN devices and
the hierarchy devices are configured to use network attack
detection or prevention techniques to identify or process malicious
traffic intended for the customer network.
16. The security system of claim 13, wherein each of the CDN
devices is configured to serve replicated content associated with
the customer network and provide security services with regard to
requests for the replicated content.
17. The security system of claim 13, wherein each of the hierarchy
devices is configured to act as a reverse proxy for traffic
intended for the customer network to obscure the presence of the
customer network.
18. A security system associated with a customer network,
comprising: a first security perimeter including a plurality of
hierarchy devices located outside of the customer network and
configured to: act as a reverse proxy for the customer network, and
determine when the customer network is the subject of a network
attack; and a second security perimeter including a plurality of
mitigation devices located outside of the customer network and
configured to: receive traffic intended for the customer network
when one of the hierarchy devices determines that the customer
network is the subject of the network attack, process the received
traffic to remove malicious traffic and leave legitimate traffic,
and forward the legitimate traffic to one or more of the hierarchy
devices.
19. The security system of claim 18, wherein a mitigation service
is initiated when the one of the hierarchy devices determines that
the customer network is the subject of the network attack, the
mitigation service causing traffic intended for the customer
network to be redirected to one or more of the mitigation
devices.
20. The security system of claim 18, wherein a tunnel is configured
between one of the mitigation devices and one of the hierarchy
devices; and wherein when forwarding the legitimate traffic to one
or more of the hierarchy devices, the one of the mitigation devices
are configured to: apply a tunnel encapsulation to the legitimate
traffic, and forward the legitimate traffic with the tunnel
encapsulation to the one of the hierarchy devices via the
tunnel.
21. The security system of claim 20, wherein the one of the
hierarchy devices is configured to: receive the legitimate traffic
with the tunnel encapsulation from the one of the mitigation
devices, remove the tunnel encapsulation, and forward the
legitimate traffic to the customer network.
22. The security system of claim 18, wherein each of the hierarchy
devices is configured to act as a gateway for traffic intended for
the customer network to obscure the presence of the customer
network.
23. A method for protecting a customer network, comprising:
receiving traffic associated with the customer network at a first
device outside the customer network; sending the traffic to a
second device outside the customer network; determining, at the
second device, whether the traffic includes malicious traffic; and
initiating a mitigation service when the traffic is determined to
include malicious traffic.
24. The method of claim 23, further comprising: determining, at the
first device, whether the traffic includes malicious traffic; and
discarding the traffic when the traffic is determined to include
malicious traffic.
25. The method of claim 24, wherein determining, at the first
device, whether the traffic includes malicious traffic includes:
determining whether the traffic includes malicious traffic based on
application-level filtering.
26. The method of claim 23, wherein sending the traffic to a second
device includes: determining, at the first device, whether the
traffic can be understood or processed, and forwarding the traffic
to the second device when the traffic cannot be understood or
processed.
27. The method of claim 23, wherein determining, at the second
device, whether the traffic includes malicious traffic includes:
determining whether the traffic includes malicious traffic based on
application-level filtering.
28. The method of claim 23, wherein initiating a mitigation service
includes: causing traffic associated with the customer network to
be directed to a third device.
29. The method of claim 28, further comprising: processing, at the
third device, the traffic to remove malicious traffic and leave
legitimate traffic; and forwarding the legitimate traffic to the
second device.
30. The method of claim 29, wherein forwarding the legitimate
traffic includes: establishing a tunnel between the third device
and the second device, and sending the legitimate traffic from the
third device to the second device via the tunnel.
Description
BACKGROUND
[0001] 1. Field of the Invention
[0002] Implementations described herein relate generally to network
security and, more particularly, to security perimeters that may be
used to protect a customer network from network attacks.
[0003] 2. Description of Related Art
[0004] Network attacks are a real, growing threat to businesses
worldwide. These attacks can quickly incapacitate a targeted
business, costing victims thousands, if not millions, of dollars in
lost revenue and productivity.
[0005] One type of network attack, called a denial of service
attack, can paralyze Internet systems by overwhelming servers,
network links, and network devices (e.g., routers, firewalls, etc.)
with bogus traffic. Easily launched against limited defenses,
network attacks not only target individual web sites or other
servers at the edge of the network, but they also can incapacitate
the network itself.
[0006] The growing dependence on the Internet makes the impact of
successful network attacks increasingly painful (financially and
otherwise) for service providers, enterprises, and government
agencies. Newer, more powerful tools used by network attackers
promise to unleash even more destructive attacks in the months and
years to come.
[0007] Because denial of service attacks are among the most
difficult to defend against, responding to them appropriately and
effectively poses a tremendous challenge for all Internet-dependent
organizations. Network devices and current perimeter security
technologies, such as firewalls and intrusion detection devices
(IDDs), although important components of an overall security
strategy, do not by themselves provide comprehensive network attack
protection.
[0008] Current techniques fall short in terms of prevention and/or
mitigation of network attacks. Some of the more popular network
attack responses, such as blackholing and router filtering, are not
optimized to deal with increasingly sophisticated attacks. IDDs
offer some excellent attack detection capabilities, but cannot
mitigate the impact of the attacks. Firewalls offer a rudimentary
level of protection but, like blackholing and router filtering,
they were not designed to protect against the types of advanced
attacks that are so common today. Still other strategies, such as
overprovisioning, do not provide adequate protection against ever
larger attacks, and can be costly as a network attack prevention
strategy.
[0009] Network attacks also make it difficult to keep up with the
constant system patching and upgrading required for a successful
security system. Many times, implementing significant patches or
upgrades is a disruptive event within a customer's network. While
overall service disruption can be avoided in a multiple server
environment, it still takes time to determine the impact of the
patch on the customer's applications and then perform the
upgrades.
SUMMARY
[0010] According to one aspect, a security system that is
associated with a customer network may include first, second, and
third security perimeters. The first security perimeter may include
a set of content delivery network (CDN) devices configured to
provide first protection against a network attack associated with
the customer network. The second security perimeter may include a
set of mitigation devices configured to provide second protection
in terms of mitigation services as a result of a network attack
associated with the customer network. The third security perimeter
may include a set of hierarchy devices configured to provide third
protection against a network attack associated with the customer
network.
[0011] According to another aspect, a security system that is
associated with a customer network may include first and second
security perimeters. The first security perimeter may include a set
of CDN devices geographically distributed in a public network. The
CDN devices may be configured to serve as a first termination point
for traffic intended for the customer network. The second security
perimeter may include a set of hierarchy devices located within the
public network. The hierarchy devices may be configured to serve as
a second termination point for traffic intended for the customer
network including traffic from the CDN devices.
[0012] According to yet another aspect, a security system that is
associated with a customer network may include first and second
security perimeters. The first security perimeter may include a set
of hierarchy devices located outside of the customer network and
configured to act as a reverse proxy for the customer network and
determine when the customer network is the subject of a network
attack. The second security perimeter may include a set of
mitigation devices located outside of the customer network and
configured to receive traffic intended for the customer network
when one of the hierarchy devices determines that the customer
network is the subject of the network attack, process the received
traffic to remove malicious traffic and leave legitimate traffic,
and forward the legitimate traffic to one or more of the hierarchy
devices.
[0013] According to a further aspect, a method for protecting a
customer network is provided. The method may include receiving
traffic associated with the customer network at a first device
outside the customer network; sending the traffic to a second
device outside the customer network; determining, at the second
device, whether the traffic includes malicious traffic; and
initiating a mitigation service when the traffic is determined to
include malicious traffic.
BRIEF DESCRIPTION OF THE DRAWINGS
[0014] The accompanying drawings, which are incorporated in and
constitute a part of this specification, illustrate one or more
embodiments of the invention and, together with the description,
explain the invention. In the drawings,
[0015] FIG. 1 is a diagram illustrating exemplary concepts
consistent with the principles of the invention for preventing and
mitigating a network attack;
[0016] FIG. 2 is a diagram of an exemplary network in which systems
and methods consistent with the principles of the invention may be
implemented;
[0017] FIG. 3 is a diagram illustrating exemplary operation of the
CDN devices of FIG. 2;
[0018] FIG. 4 is a diagram illustrating exemplary operation of the
mitigation devices of FIG. 2;
[0019] FIG. 5 is a diagram illustrating exemplary operation of the
hierarchy devices of FIG. 2;
[0020] FIG. 6 is a flowchart of exemplary processing for providing
security against network attacks according to an implementation
consistent with the principles of the invention;
[0021] FIGS. 7A and 7B are diagrams illustrating exemplary
operation of the CDN, mitigation, and hierarchy devices according
to an implementation consistent with the principles of the
invention;
[0022] FIGS. 8-10 are diagrams of exemplary combinations of
security perimeters that may be used in other implementations
consistent with the principles of the invention; and
[0023] FIG. 11 is a flowchart of an exemplary implementation
consistent with the principles of the invention.
DETAILED DESCRIPTION
[0024] The following detailed description of the invention refers
to the accompanying drawings. The same reference numbers in
different drawings may identify the same or similar elements. Also,
the following detailed description does not limit the
invention.
Overview
[0025] Network attacks, such as distributed denial of service
(DDoS) attacks, work by taking advantage of Internet protocols and
the fundamental benefit of the Internet of delivering datagrams
from nearly any source to any destination, without prejudice.
Essentially, it is the behavior of these datagrams that defines the
network attack: either there are too many, overwhelming network
devices as well as servers, or they are deliberately incomplete to
rapidly consume server resources. What makes network attacks so
difficult to prevent is that illegitimate datagrams are
indistinguishable from legitimate datagrams, making detection
difficult. Many of these attacks also use spoofed source IP
addresses, thereby eluding source identification by anomaly-based
monitoring tools looking for unusually high volumes of traffic
coming from specific origins.
[0026] A growing trend among network attackers is to use
sophisticated spoofing techniques and essential protocols (instead
of nonessential protocols that can be blocked) to make network
attacks even more stealthy and disruptive. These attacks, which use
legitimate application protocols and services, are very difficult
to identify and defeat. Employing datagram-filtering or
rate-limiting measures simply completes the attacker's task by
shutting everything down, causing denial of legitimate traffic.
[0027] FIG. 1 is a diagram illustrating exemplary concepts
consistent with the principles of the invention for preventing and
mitigating a network attack. FIG. 1 shows a customer network 110
and three security perimeters (e.g., CDN perimeter 140, mitigation
perimeter 160, and hierarchy perimeter 180). The three security
perimeters are shown located on the public side of the connection
of customer network 110 to a public network 120, such as the
Internet.
[0028] Customer network 110 may include hardware and software
systems associated with a customer's infrastructure. Customer
network 110 may include security systems to protect the customer's
infrastructure from network attacks. In one implementation,
customer network 110 may provide a service by which it makes
content available on public network 120.
[0029] CDN perimeter 140 may include a group of CDN devices 142 to
provide a first level of protection against network attacks. The
particular number of CDN devices 142 used in CDN perimeter 140 may
be a design decision. A CDN has traditionally been used to deliver
web content to a user based on the geographic locations of the
user, the origin of the web content, and its CDN devices. The CDN
copies the content of a web server, for example, to a network of
CDN devices that are dispersed at geographically different
locations. When a user requests content that is part of a CDN, the
CDN will direct the request to a CDN device that is closest to the
user to deliver the content to the user. In one implementation, CDN
perimeter 140 may provide a range of security services on CDN
devices 142 for the specific applications that can be served by a
CDN device 142. For these applications, CDN device 142 can offer
protection against network attacks, core site obscurement,
application level filtering, and protection from disruptive
operating system and application patching and upgrading.
[0030] Mitigation perimeter 160 may include a group of mitigation
devices 162 to provide mitigation services, as a second level of
protection, in the event of a network attack. The particular number
of mitigation devices 162 used in mitigation perimeter 160 may be a
design decision. When customer network 110 discovers that it is
under attack, traffic intended for customer network 110 can be
diverted to a mitigation device, such as one of mitigation devices
162. Mitigation device 162 may process the traffic to remove all
malicious traffic while allowing good traffic to continue flowing
uninterrupted. After the traffic has been processed to remove
malicious traffic, mitigation device 162 may forward the good
traffic to customer network 110 or hierarchy perimeter 180.
[0031] Hierarchy perimeter 180 may include a group of hierarchy
devices 182 to provide a third level of protection against network
attacks. The particular number of hierarchy devices 182 used in
hierarchy perimeter 180 may be a design decision. Hierarchy devices
182 may interface with customer network 110 and act as reverse
proxies. A reverse proxy acts as a gateway to a customer's network
by acting as the final IP address for requests from outside the
customer's network. From outside the customer's network, the
reverse proxy appears as a system within the customer's network.
Similarly, hierarchy devices 182 may act as final IP addresses for
systems of customer network 110.
[0032] CDN perimeter 140, mitigation perimeter 160, and hierarchy
perimeter 180 may expand the security system for customer network
110 well beyond the connection point of customer network 110 to
public network 120. Combining these perimeters into the security
system of customer network 110 permits customer network 110 to
withstand nearly any scale network attack, transparently increase
its serving capacity, make its network invisible to public network
120, and patch and upgrade its systems on its own schedule and not
the schedule of the latest vulnerability.
Exemplary Network Configuration
[0033] FIG. 2 is a diagram of an exemplary network 200 in which
systems and methods consistent with the principles of the invention
may be implemented. Network 200 may include customer network 110
connected to public network 120. The particular systems making up
customer network 110 may vary. As shown in FIG. 2, for example,
customer network 110 may include a group of customer devices 210
connected to an IDD 220 and a firewall 230 via a network 240. In
other implementations, customer network 110 may include more,
fewer, or different systems. Also, in some instances, two or more
of the systems may be implemented as a single device and/or one of
the systems may be implemented as two or more (possibly
distributed) devices.
[0034] Customer devices 210 may include personal computers,
workstations, personal digital assistants (PDAs), lap tops,
wireless or landline telephone systems, and/or other types of
computation or communication devices. In one implementation, one or
more of customer devices 210 may operate as server devices that
store and/or manage information and make at least some of this
information available to one or more client devices outside of
customer network 110.
[0035] IDD 220 may include a security device that inspects inbound
(and possibly outbound) network activity and identifies suspicious
patterns that may indicate an attack on customer network 110. IDD
220 may provide misuse detection by which IDD 220 analyzes the
information it gathers and compares it to a large database of
attack signatures. Alternatively, or additionally, IDD 220 may
provide anomaly detection by which IDD 220 monitors network
segments to compare their state to a normal baseline of network
traffic load, breakdown, protocol, and typical packet size and
identifies any anomalies. IDD 220 may operate in passive and/or
reactive modes. In the passive mode, IDD 220 may detect a potential
security breach, log the information, and signal an alert. In the
reactive mode, IDD 220 may respond to suspicious activity by
logging off a user or by reprogramming firewall 230 to block
network traffic from the suspected malicious source.
[0036] Firewall 230 may include hardware and/or software that
prevents unauthorized access to or from customer network 110.
Firewall 230 may examine each message entering or leaving customer
network 110 and block those that do not meet specified security
criteria.
[0037] Network 240 may include a local area network (LAN), a wide
area network (WAN), a private network, an intranet, or a
combination of networks. Customer devices 210, IDD 220, and
firewall 230 may connect to network 240 via wired, wireless, and/or
optical connections.
[0038] Public network 120 may include a WAN, a telephone network,
such as the Public Switched Telephone Network (PSTN), an intranet,
the Internet, or a combination of networks. Various good and bad
sources of traffic may connect to public network 120 via wired,
wireless, and/or optical connections. An example of a good source
of traffic may include a legitimate web surfer. An example of a bad
source of traffic may include a source of a DDoS attack.
[0039] As described with regard to FIG. 1, public network 120 may
include three security perimeters. For example, public network 120
may include a group of CDN devices 142 (associated with the CDN
perimeter), a group of mitigation devices (MD) 162 (associated with
the mitigation perimeter), and a group of hierarchy devices (HD)
182 (associated with the hierarchy perimeter). Each of these
devices 142, 162, and 182 may operate alone and in combination to
provide additional levels of security for customer network 110.
Accordingly, each of devices 142, 162, and 182 will be described in
terms of the security features they offer alone and then how they
work together to provide superior security for customer network
110.
Security Provided By Cdn Perimeter
[0040] The CDN perimeter may make up the outermost layer of a
security system according to an implementation consistent with the
principles of the invention. The CDN perimeter makes it possible to
obscure the presence of customer network 110 from the standpoint of
public network 120, while increasing the scalability and
reliability of customer network 110.
[0041] During the evolution of CDNs, the nature of the service
paved the way for a number of security functions. Traditionally
thought of as simply "technology making web-serving faster," the
CDN enables a range of security services for the specific
applications that can be served by the CDN. These applications
might include, for example, HTTP-based applications, HTTPS-based
applications, FTP services, and streaming media services.
[0042] FIG. 3 is a diagram illustrating exemplary operation of CDN
devices 142. CDN devices 142 may include conventional CDN devices
that are modified to include security features that may be used on
behalf of customer network 110. CDN devices 142 may be dispersed at
geographically different locations.
[0043] Each CDN device 142 may appear as a discrete service
termination point for customer network 110. Therefore, the traffic
handled by CDN devices 142 does not reach customer network 110,
thereby obscuring the presence of customer network 110. In some
situations, a CDN device 142 may contact customer network 110, such
as when needing content to satisfy a user's request or when passing
on a request that it does not understand.
[0044] By design, different end users are provided with different
IP addresses (associated with particular ones of CDN devices 142)
based on geographic location. To bad sources (i.e., sources of
malicious traffic, such as DDoS attack servers), customer network
110 appears to be in various locations. Because malicious traffic
from bad sources may originate from different geographic locations,
each individual bad source may end up attacking a different CDN
device 142. This serves to diffuse the attack. As a result, a
single CDN device 142 needs not bear the brunt of the attack. In
the event that bad sources are configured to attack one specific
CDN device 142 by hard-coded IP, for example, that CDN device 142
may be configured to send legitimate traffic to other CDN devices
142 providing the service.
[0045] The protection against attack may extend beyond the initial
diffusion of the attack. Each CDN device 142 may be hardened
against attack, using any well known network attack
detection/prevention techniques to thwart known attacks. Each CDN
device 142 may be specialized and tuned for maximum performance and
resilience in the face of attack. For example, each CDN device 142
may include a set of application-level rules that it may use to
filter traffic for security purposes. The rules may identify the
types of traffic and/or applications that are known to be capable
of exploiting a vulnerability. When a CDN device 142 receives
traffic identified by the rules or associated with an application
identified by the rules, it may discard the traffic.
[0046] As explained above, CDN devices 142 may act as termination
points for certain traffic for customer network 110 and, thereby,
obscure the presence of customer network 110.
[0047] Using a CDN perimeter to protect the infrastructure of a
network creates a number of opportunities for cost savings. One of
the impacts of using a CDN perimeter is bandwidth offloading. This
is accomplished using standard CDN technologies to serve various
types of objects to end users directly. That bandwidth no longer
crosses any of the servers or circuits in the infrastructure of
customer network 110. This reduction in bandwidth is also
accompanied by a reduction in processor use across the public
facing infrastructure of customer network 110. Some processor
intensive operations, such as encryption of objects for secure
traffic, may be performed by CDN devices 142.
[0048] A reduction in bandwidth and processor use can be directly
translated into a reduction of infrastructure required. Less
circuit capacity to public network 120 may be needed, and a
multi-server infrastructure can see reclamation of servers, which
can be re-tasked for other pending projects. Planned upgrades to
expand delivery capacity can be moved out into the future or
eliminated entirely. Finally, the CDN security protection extends
beyond malicious traffic. An unexpected but legitimate high volume
event may have little impact on the infrastructure of customer
network 110 since CDN devices 142 will handle the vast majority of
the load.
Security Provided By Mitigation Perimeter
[0049] Network attackers typically engage a wide variety of sources
to target a customer's network. While CDN devices 142 can obscure
the presence of customer network 110, an attack can still be
launched directly using the IP addresses associated with systems of
customer network 110. The mitigation perimeter may offer another
level of security in the form of mitigation services.
[0050] FIG. 4 is a diagram illustrating exemplary operation of
mitigation devices 162. Mitigation devices 162 may include devices
capable of scrubbing traffic to remove malicious traffic while
allowing good traffic to flow uninterrupted. In one implementation,
a mitigation device 162 may subject traffic to a multi-verification
process that entails dynamic filtering, activation verification,
anomaly recognition, protocol analysis, and rate limiting. In
another implementation, a mitigation device 162 may include some
other known type of device that is capable of scrubbing traffic to
remove malicious traffic.
[0051] In one implementation, mitigation devices 162 may be
dispersed at geographically different locations. Optionally, sets
of mitigation devices 162 may be formed into mitigation centers.
Mitigation devices 162 in the different mitigation centers may
offer redundancy. Therefore, if a mitigation device 162 within a
mitigation center or the mitigation center itself fails, another
mitigation device 162 or mitigation center can be used.
[0052] When an attack, such as a DDoS attack, is detected at
customer network 110, a mitigation service may be initiated. The
mitigation service may redirect traffic intended for customer
network 110 to one (or more) of mitigation devices 162. Mitigation
device(s) 162 may scrub the traffic to remove malicious traffic and
leave the good traffic.
[0053] Mitigation device(s) 162 may forward the good traffic to
customer network 110 using a tunnel, such as a generic routing
encapsulation (GRE) tunnel. A GRE tunnel may be used to traverse
traffic from one place in a network to another place in the network
without requiring a lot of routing decisions to be made in between
those two places. To forward the good traffic to customer network
110, mitigation device(s) 162 may apply a tunnel encapsulation to
the traffic. For example, mitigation device(s) 162 may add another
header to the traffic that includes both the source address (i.e.,
the address of mitigation device 162) and the destination address
(i.e., an address associated with customer network 110). A network
device that receives the tunnel traffic may recognize it as tunnel
traffic and forward it onto its destination (i.e., customer network
110). When customer network 110 receives the tunnel traffic, it may
strip the encapsulation and process or forward the traffic, as
necessary.
Security Provided By Hierarchy Perimeter
[0054] Network attackers typically engage a wide variety of sources
to target a customer's network. While CDN devices 142 can obscure
the presence of customer network 110 to certain types of
applications, other types of applications may be used to launch an
attack directly at customer network 110. The hierarchy perimeter
may offer another level of security to further obscure the presence
of customer network 110.
[0055] FIG. 5 is a diagram illustrating exemplary operation of
hierarchy devices 182. Hierarchy devices 182 may include devices
capable of acting as reverse proxies on behalf of customer network
110. For example, hierarchy devices 182 may act as a gateway to
customer network 110 by acting as the final IP address for traffic
destined for customer network 110.
[0056] Hierarchy devices 182 may be configured to receive all
traffic intended for customer network 110, including traffic from
CDN devices 142 (not shown), traffic from applications not handled
by CDN devices 142, and traffic from mitigation devices 162 (not
shown). Each hierarchy device 182 may be hardened against attack,
using any well known network attack detection/prevention techniques
to thwart known attacks. For example, each hierarchy device 182 may
include a set of application-level rules that it may use to filter
traffic for security purposes. The rules may identify the types of
traffic and/or applications that are known to be capable of
exploiting a vulnerability. When a hierarchy device 182 receives
traffic identified by the rules or associated with an application
identified by the rules, it may discard the traffic.
[0057] By instituting this type of hierarchy, customer network 110
can rigidly wall off its core infrastructure from the outside
world. Customer network 110 can, thereby, limit all traffic flows
reaching its firewall to the set of hierarchy devices 182. Even if
an IP address associated with a system of customer network 110 was
somehow discovered by a would-be attacker, hierarchy devices 182
may block the malicious traffic before it reached the outside of
the firewall.
[0058] Because hierarchy devices 162 act as a gateway for traffic
intended for customer network 110, they obscure the presence of
customer network 110 from the standpoint of public network 120. As
a result, operators of customer network 110 may perform system
patching and upgrading (e.g., security patching and upgrading) in a
more organized and less disruptive fashion.
Exemplary Processing
[0059] FIG. 6 is a flowchart of exemplary processing for providing
security against network attacks according to an implementation
consistent with the principles of the invention. Processing may
include providing a CDN perimeter, a mitigation perimeter, and a
hierarchy perimeter for a customer network, such as customer
network 110 (blocks 610-630). The CDN perimeter may include a group
of CDN devices 142 that are geographically distributed in a public
network, such as public network 120. The mitigation perimeter may
include a group of mitigation devices 162, possibly formed into
mitigation centers and geographically distributed in public network
120. The hierarchy perimeter may include a group of hierarchy
devices 182 within public network 120. The particular number of CDN
devices 142, mitigation devices 162, and/or hierarchy devices 182
within these perimeters may be determined based upon one or more
parameters, such as cost, estimated bandwidth of traffic for
customer network 110, etc.
[0060] The CDN perimeter, mitigation perimeter, and hierarchy
perimeter may be configured to operate on behalf of customer
network 110 (block 640). For example, CDN devices 142 may be
configured as termination points for certain traffic intended for
customer network 110. CDN devices 142 may also be configured to
direct its own traffic (e.g., CDN traffic) to hierarchy devices
182. Mitigation devices 162 may be configured to process traffic
intended for customer network 110 upon initiation of a mitigation
service. A tunnel, such as GRE tunnel, may be established
beforehand between mitigation devices 162 and hierarchy devices
182. Mitigation devices 162 may use the tunnel to send legitimate
traffic to hierarchy devices 182. Hierarchy devices 182 may be
configured as the gateway for all traffic intended for customer
network 110. As such, hierarchy devices 182 may be configured as
termination points for traffic intended for customer network
110.
[0061] The CDN perimeter, mitigation perimeter, and hierarchy
perimeter may be used to secure customer network 110 from network
attacks (block 650). For example, CDN devices 142 may be configured
as a first layer of security against certain network attacks.
Mitigation devices 162 may provide mitigation services in the event
of a network attack. Hierarchy devices 182 may provide a final
layer of security against network attacks prior to reaching
customer network 110.
[0062] FIGS. 7A and 7B are diagrams illustrating exemplary
operation of the CDN, mitigation, and hierarchy devices according
to an implementation consistent with the principles of the
invention. While FIGS. 7A and 7B show two CDN devices 142, one
mitigation device 162, and two hierarchy devices 182, there may be
more or fewer of these devices in practice.
[0063] As shown in FIG. 7A, there are several ways that a traffic
source might interface with customer network 110. For example, good
sources 710 and 720 and bad source(s) 730 might communicate with
CDN devices 142. Good sources 740 and 750 and bad source(s) 760
might communicate with hierarchy devices 182. It might also be
possible (though unlikely) for a source, such as bad source(s) 770,
to bypass the security perimeters and communicate directly with
customer network 110.
[0064] With regard to good sources 710 and 720, CDN devices 142 may
receive legitimate (good) traffic from and send traffic to these
sources. For example, CDN devices 142 may receive traffic in the
form of requests for content. CDN devices 142 may use one or more
well known network attack detection/prevention techniques to
determine whether the requests are legitimate or malicious. A CDN
device 142 may process a legitimate one of these requests by
providing the requested content, obtaining the content or related
content from customer network 110, directing the request to another
CDN device 142, or, when CDN device 142 does not understand or
cannot process the request, sending the request to customer network
110 via hierarchy device 182.
[0065] With regard to bad source(s) 730, CDN device 142 may receive
malicious (bad) traffic from bad source(s) 730. CDN device 142 may
use one or more well known network attack detection/prevention
techniques, such as application-level filtering, to detect and
discard the malicious traffic. For example, CDN device 142 may
include a list of known malicious traffic and/or applications. When
CDN device 142 receives traffic identified on its list or
associated with an application identified on its list, it may
discard the traffic.
[0066] It may be possible for bad source(s) 730 to provide
malicious traffic that is not recognized by CDN device 142. In this
case, CDN device 142 may forward the malicious traffic to a
hierarchy device 182. Hierarchy device 182 may use one or more well
known network attack detection/prevention techniques, such as
application-level filtering or filtering of finer granularity, to
detect and discard the malicious traffic. For example, hierarchy
device 182 may include a list of known malicious traffic and/or
applications. When hierarchy device 182 receives traffic identified
on its list or associated with an application identified on its
list, it may discard the traffic.
[0067] With regard to good sources 740 and 750, hierarchy devices
182 may receive legitimate (good) traffic from and send traffic to
these sources. Hierarchy devices 182 may use one or more well known
network attack detection/prevention techniques, such as
application-level filtering or filtering of finer granularity, to
determine whether traffic from these sources is legitimate or
malicious. Hierarchy devices 182 may forward legitimate traffic
onto customer network 110.
[0068] With regard to bad source(s) 760, hierarchy device 182 may
receive malicious (bad) traffic from bad source(s) 760. Hierarchy
device 182 may use one or more well known network attack
detection/prevention techniques, such as application-level
filtering or filtering of finer granularity, to detect and discard
the malicious traffic. For example, hierarchy device 182 may
include a list of known malicious traffic and/or applications. When
hierarchy device 182 receives traffic identified on its list or
associated with an application identified on its list, it may
discard the traffic.
[0069] With regard to bad source(s) 770, it may be possible (though
unlikely) for a network attacker to direct an attack directly on
customer network 110. This may occur, for example, when the network
attacker somehow discovers an IP address of a system of customer
network 110. When this occurs, the firewall of customer network 110
may determine that customer network 110 is the subject of a network
attack.
[0070] When customer network 110 is the subject of a network
attack, the mitigation service may be initiated. There are a few
ways that the mitigation service may be initiated. For example,
customer network 110 (or a person associated with customer network
110) may initiate the mitigation service by notifying a network
device, such as a hierarchy device 182 or a gateway associated with
customer device 110, or a security operations center (SOC).
Alternatively, the mitigation service may be initiated without
input from customer network 110. For example, hierarchy device 182
or a SOC may detect that customer network 110 is the subject of an
attack and may initiate the mitigation service on behalf of
customer network 110.
[0071] In either event, the next-hop route associated with customer
network 110 may be set to mitigation device 162. This next-hop
route may be advertised on public network 120. For example, the
next-hop route for customer network 110 may be sent to other
devices in public network 120. These other devices may update their
routing tables based on the next-hop route.
[0072] As shown in FIG. 7B, customer traffic may then be redirected
to mitigation device 162. The customer traffic may include
datagrams from one or more of good and/or bad sources 710-770
and/or datagrams from CDN devices 142. When a network device within
public network 120 receives a datagram destined for customer
network 110, the network device may identify mitigation device 162
as the destination for the datagram. When there are multiple
mitigation devices 162, the network device may select the
appropriate one of mitigation devices 162 to receive the
datagram.
[0073] Mitigation device 162 may scrub the traffic to remove
malicious traffic and leave the good traffic. In one
implementation, mitigation device 162 may subject the traffic to a
multi-verification process that entails dynamic filtering,
activation verification, anomaly recognition, protocol analysis,
and rate limiting. In another implementation, mitigation device 162
may use another technique to remove malicious traffic.
[0074] Mitigation device 162 may forward the good traffic to a
hierarchy device 182 using a tunnel, such as a GRE. As explained
previously, a GRE tunnel may be used to traverse traffic from one
place in a network to another place in the network without
requiring a lot of routing decisions to be made in between those
two places. In this case, a GRE tunnel may be established between
mitigation device 162 and hierarchy device 182.
[0075] To forward the good traffic to hierarchy device 182,
mitigation device 162 may apply a tunnel encapsulation to the
traffic. For example, mitigation device 162 may add another header
to the traffic that includes both the source address (i.e., the
address of mitigation device 162) and the destination address
(i.e., the address associated with hierarchy device 182). A network
device that receives the tunnel traffic may recognize it as tunnel
traffic and forward it onto its destination (i.e., hierarchy device
182). When hierarchy device 182 receives the tunnel traffic, it may
strip the encapsulation and process and forward the traffic, as
necessary, to customer network 110.
[0076] The additional security that the above-described security
perimeters provide to customer network 110 permit system patches
and upgrades (e.g., security patches and upgrades) to be performed
within customer network 110 in a more organized and less disruptive
fashion. One of the most difficult aspects of securing a customer's
infrastructure is keeping up with the constant system patching and
upgrading required. Many times, implementing significant patches is
a disruptive event within the customer's infrastructure. While
overall service disruption can be avoided in a multiple server
environment, it still takes time to determine the impact of the
patch on the customer's applications and then perform the upgrades.
With the CDN, mitigation, and hierarchy perimeters in place,
customer network 110 may patch and upgrade on its own schedule.
[0077] According to implementations consistent with the principles
of the invention, CDN devices 142 and hierarchy devices 182 may
appear as the termination points for customer network 110, and
hierarchy devices 182 may be the only devices communicating
directly with customer network 110. A particular type of request
that might exploit the vulnerability of customer network 110 and
breach its systems can be easily blocked at the edge of the CDN or
hierarchy perimeters. The hierarchy perimeter ensures that a
network attacker cannot bypass the CDN perimeter and issue damaging
requests directly to customer network 110. Any type of request
known to be capable of exploiting a vulnerability can be listed for
CDN devices 142 and/or hierarchy devices 182 to ignore.
[0078] Some vulnerability patches can create problems with third
party or custom coded applications. With the knowledge that the new
security perimeters are blocking malicious application level
requests, customer network 110 can perform normal integration and
regression testing of a new patch in its staging environment(s).
This dramatically reduces the chance that a "quick security patch"
does not inadvertently cause significant service failures/crashes.
Security patch scheduling can also be aligned with normal
development release cycles and not waste cycles testing patches
separately.
Combinations Of Security Perimeters
[0079] While three security perimeters have been described as
inter-operating to provide security to customer network 110, other
combinations of these security perimeters may be used in other
implementations consistent with the principles of the invention.
FIGS. 8-10 are diagrams of exemplary combinations of security
perimeters that may be used in other implementations consistent
with the principles of the invention. As shown in FIG. 8, the CDN
and hierarchy perimeters may operate together to provide security
to customer network 110. As shown in FIG. 9, the mitigation and
hierarchy perimeters may operate together to provide security to
customer network 110. As shown in FIG. 10, the CDN and mitigation
perimeters may operate together to provide security to customer
network 110. The inter-operation of these perimeters should be
generally understood based on the description provided
previously.
Exemplary Implementation
[0080] FIG. 11 is a flowchart of an exemplary implementation
consistent with the principles of the invention. The processing of
FIG. 11 may begin with reception of traffic associated with a
customer network at a first device that may be located outside of
the customer network (block 1110). In one implementation, the first
device may include a CDN device. The first device may determine
whether the traffic can be understood or processed.
[0081] When the traffic can be understood or processed, the first
device may determine whether the traffic includes malicious
traffic. The first device may use one or more well known network
attack detection/prevention techniques to determine whether the
traffic is legitimate or malicious. When the traffic is legitimate,
the first device may process the traffic as described above. When
the traffic is malicious, the first device may discard the traffic
or take some other known security measure.
[0082] When the traffic cannot be understood or processed, the
first device may send the traffic to a second device that may also
be located outside of the customer network (block 1120). In one
implementation, the second device may include a hierarchy device.
The second device may determine whether the traffic includes
malicious traffic (block 1130). The second device may use one or
more well known network attack detection/prevention techniques to
determine whether the traffic is legitimate or malicious. If the
traffic is legitimate, then the second device may process the
traffic, if necessary, and/or forward the traffic to the customer
network.
[0083] When the traffic is malicious, a mitigation service may be
initiated (block 1140). In this case, customer traffic may be
redirected to a third device. In one implementation, the third
device may include a mitigation device. The third device may
process the traffic to remove malicious traffic and leave
legitimate traffic. The third device may forward the legitimate
traffic to the second device via, for example, a tunnel.
Conclusion
[0084] Systems and methods consistent with the principles of the
invention may provide multiple security perimeters to secure a
customer's network. These security perimeters may obscure the
presence of the customer's network and permit the customer to
perform system patches and upgrades in a more organized and less
disruptive fashion.
[0085] The foregoing description of preferred embodiments of the
present invention provides illustration and description, but is not
intended to be exhaustive or to limit the invention to the precise
form disclosed. Modifications and variations are possible in light
of the above teachings or may be acquired from practice of the
invention.
[0086] For example, while series of acts have been described with
regard to FIGS. 6 and 11, the order of the acts may be modified in
other implementations consistent with the principles of the
invention. Further, non-dependent acts may be performed in
parallel.
[0087] Also, it has been described that various devices transmit
and receive network traffic. The term "traffic" is intended to
refer to any type or form of data, including packet and non-packet
data. The term "datagram" may also be used to refer to any type or
form of data, such as packet and non-packet data.
[0088] It will be apparent to one of ordinary skill in the art that
aspects of the invention, as described above, may be implemented in
many different forms of software, firmware, and hardware in the
implementations illustrated in the figures. The actual software
code or specialized control hardware used to implement aspects
consistent with the principles of the invention is not limiting of
the invention. Thus, the operation and behavior of the aspects were
described without reference to the specific software code it being
understood that one of ordinary skill in the art would be able to
design software and control hardware to implement the aspects based
on the description herein.
[0089] No element, act, or instruction used in the present
application should be construed as critical or essential to the
invention unless explicitly described as such. Also, as used
herein, the article "a" is intended to include one or more items.
Where only one item is intended, the term "one" or similar language
is used. Further, the phrase "based on" is intended to mean "based,
at least in part, on" unless explicitly stated otherwise.
* * * * *